RE: Some Draft SchemaBibEx Proposals - string identifier with no URI

Note that URNs are URIs. URNs never fulfilled their promise for several reasons:
1) Unlike http URIs, they aren't backed by a protocol
2) IETF was lax about registering additional URN namespaces
3) An effort to delegate URI namespace assignments collapsed ("info" URIs)
4) Web standards were updated to eliminate limitations on http URIs:
 - http://tools.ietf.org/html/rfc3305

 - http://www.w3.org/2001/tag/group/track/issues/14


The fate of "info" URIs illustrates the point:

http://info-uri.info/


In a nutshell, http URIs are infinitely better than other forms of identifiers. Note the 1st two principles of Linked Data:

1) Use URIs as names for things
2) Use HTTP URIs so that people can look up those names

http URIs are the identifiers of the future. The way to upgrade our legacy "string" identifiers is for the agency to prefix them with the letters "http://" ASAP. The old string identifiers can still be related using the mechanisms being described, but they will grow increasingly marginal.

Jeff

> -----Original Message-----
> From: Shlomo Sanders [mailto:Shlomo.Sanders@exlibrisgroup.com]
> Sent: Wednesday, January 16, 2013 12:51 PM
> To: Laura Dawson; Wallis,Richard; Karen Coyle
> Cc: public-schemabibex@w3.org
> Subject: Re: Some Draft SchemaBibEx Proposals - string identifier with
> no URI
> 
> Can we get back what to about an identifier like urn:nbn:ch:bel-9039
> where there is no URI?
> 
> Thanks,
> Shlomo
> 
> Sent from my iPad
> 
> On Jan 16, 2013, at 19:37, "Laura Dawson" <ljndawson@gmail.com> wrote:
> 
> > Seems like it would be pretty easy to set this up. I'll put it on the
> > roadmap - and ignore the lectures from the development team about how
> > I'm not paying attention to revenue-producing projects.
> >
> > On 1/16/13 12:27 PM, "Richard Wallis" <richard.wallis@oclc.org>
> wrote:
> >
> >> Description of an isbn:
> >>   name (characters that make up the number): 093738318X
> >>   type (the numbering scheme it is in): ISBN
> >>   issuing authority: Bowker
> >>   issue date: 1997
> >>   assigned to: http://www.bookwire.com/9780985887025

> >>
> >> OR using the SKOS proposal:
> >>
> >> <http://bowker.com/identifiers/isbn/9780553479430>
> >>   a skos:Concept;
> >>   schema:name "093738318X";
> >>   schema:inScheme <http://bowker.com/concept-scheme/isbn> ;
> >>   schema:focus <http://www.bookwire.com/9780985887025>.
> >>
> >> Markup of the concept-scheme could, in this model, provide
> >> information about the issuing authority.  Not sure yet how I would
> >> map the issue date property
> >>
> >> ~Richard.
> >>
> >> On 16/01/2013 17:14, "Laura Dawson" <ljndawson@gmail.com> wrote:
> >>
> >>> Good point.
> >>>
> >>> When you say "description of the number" - could you give an
> example?
> >>>
> >>> On 1/16/13 12:03 PM, "Richard Wallis" <richard.wallis@oclc.org>
> wrote:
> >>>
> >>>> Is the Bookwire page not a description of the book (that happens
> to
> >>>> have that isbn allocated to it) not a description of the number
> >>>> itself?
> >>>> ~Richard.
> >>>>
> >>>> On 16/01/2013 16:56, "Laura Dawson" <ljndawson@gmail.com> wrote:
> >>>>
> >>>>> We do, actually - we just make a different URL available publicly
> >>>>> for SEO purposes. It's not on Bowker.com - it's on a site called
> >>>>> Bookwire.
> >>>>>
> >>>>> http://www.bookwire.com/9780985887025 is an example. It's an
> alias
> >>>>> for http://www.bookwire.com/The-Twelve-9780985887025.html.

> >>>>>
> >>>>>
> >>>>> On 1/16/13 11:49 AM, "Karen Coyle" <kcoyle@kcoyle.net> wrote:
> >>>>>
> >>>>>> I admit that I always have trouble with the re-working of
> >>>>>> ontologies to new uses (like using dcterms:title for a person's
> >>>>>> name, which is legit but always rubs me the wrong way). In the
> >>>>>> SKOS case, I just can't see an identifier as a skos:concept.
> >>>>>> Also, if Bowker *did* provide a URI for ISBNs (and I think
> that's
> >>>>>> being discussed but is not yet realized) then I see no need for
> >>>>>> the identifier structure in schema. It is needed for those
> >>>>>> instances where there is no URI. (But, Jeff, maybe that's just
> an
> >>>>>> artifact of your example?)
> >>>>>>
> >>>>>> kc
> >>>>>>
> >>>>>> On 1/15/13 11:07 PM, Shlomo Sanders wrote:
> >>>>>>> I looked at both. They seem to be equivalent with the SKOS
> being
> >>>>>>> cleaner and also based on only one new construct: SKOS.
> >>>>>>> Personally, I think the name attribute in the SKOS is a
> >>>>>>> misleading attribute label for at text key value.
> >>>>>>> If I had to choose between name and prefLabel I would prefer
> the
> >>>>>>> latter (though that is also misleading too but better than
> >>>>>>> name).
> >>>>>>> How does this work for?
> >>>>>>> schema:identifier <urn:nbn:ch:bel-9039>; In the SKOS part
> >>>>>>> itself, is the do the inSchema and focus need to be working
> URIs
> >>>>>>> with something behind it?
> >>>>>>> <http://bowker.com/identifiers/isbn/9780553479430>
> >>>>>>>     a skos:Concept;
> >>>>>>>     schema:name "9780553479430";
> >>>>>>>     schema:inScheme <http://bowker.com/concept-scheme/isbn> ;
> >>>>>>>     schema:focus <http://www.worldcat.org/oclc/38264520>.
> >>>>>>> Shlomo
> >>>>>>> -----Original Message-----
> >>>>>>> From: Young,Jeff (OR) [mailto:jyoung@oclc.org]
> >>>>>>> Sent: Tuesday, January 15, 2013 20:05
> >>>>>>> To: Shlomo Sanders; Wallis,Richard; Karen Coyle; Gordon Dunsire
> >>>>>>> Cc: public-schemabibex@w3.org; Vizine-Goetz,Diane
> >>>>>>> Subject: RE: Some Draft SchemaBibEx Proposals There was some
> >>>>>>> discussion of Richard's Identifier Proposal during today's
> call,
> >>>>>>> so I wanted to clarify my comments.
> >>>>>>> My observation was that the key patterns in Richard's
> >>>>>>> "Identifier Proposal" mirror patterns found in SKOS. I added a
> >>>>>>> section to the proposal so they can be compared side-by-side:
> >>>>>>>
> http://www.w3.org/community/schemabibex/wiki/Identifier#Proposal
> >>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>>
> http://www.w3.org/community/schemabibex/wiki/Identifier#Alternat

> >>>>>>> e_Pro
> >>>>>>> po
> >>>>>>> se
> >>>>>>> d_based_on_SKOS
> >>>>>>> If Schema.org adopts the essence of SKOS (which they should if
> >>>>>>> they're serious about wanting to externalize lists), then
> >>>>>>> Richard's Proposal can be modeled as a specialization of that:
> >>>>>>> schema:Identifier rdfs:subClassOf schema:Concept .
> >>>>>>> schema:inStandard rdfs:subPropertyOf schema:inScheme .
> >>>>>>> schema:identifies rdfs:subPropertyOf schema:focus .
> >>>>>>> I'm skeptical that Schema.org will care about explicitly
> >>>>>>> modeling "identifiers for identifiers" like this, but I won't
> >>>>>>> object if the group wants to try.
> >>>>>>> Jeff
> >>>>>>>> -----Original Message-----
> >>>>>>>> From: Shlomo Sanders [mailto:Shlomo.Sanders@exlibrisgroup.com]
> >>>>>>>> Sent: Monday, January 07, 2013 1:30 AM
> >>>>>>>> To: Wallis,Richard; Karen Coyle; Gordon Dunsire
> >>>>>>>> Cc: public-schemabibex@w3.org; Young,Jeff (OR);
> >>>>>>>> Vizine-Goetz,Diane
> >>>>>>>> Subject: RE: Some Draft SchemaBibEx Proposals
> >>>>>>>>
> >>>>>>>> "So, in my draft I was intending the URI would identify (link
> >>>>>>>> to) the description of a [standard] identifier."
> >>>>>>>>
> >>>>>>>> This seems convoluted and not KISS.
> >>>>>>>>
> >>>>>>>> -----Original Message-----
> >>>>>>>> From: Richard Wallis [mailto:richard.wallis@oclc.org]
> >>>>>>>> Sent: Sunday, January 06, 2013 22:12
> >>>>>>>> To: Karen Coyle; Gordon Dunsire
> >>>>>>>> Cc: public-schemabibex@w3.org; 'Young,Jeff (OR)';
> >>>>>>>> Vizine-Goetz,Diane
> >>>>>>>> Subject: Re: Some Draft SchemaBibEx Proposals
> >>>>>>>>
> >>>>>>>> Hi Gordon,
> >>>>>>>>
> >>>>>>>> As Karen mentions, Schema.org purposely avoids the strict
> >>>>>>>> definition of domains and ranges and (as I would put it)
> >>>>>>>> 'hopes' to find the Expected Type as a property but it is also
> >>>>>>>> acceptable to find a string representation.
> >>>>>>>>
> >>>>>>>> The 'identifier' property I describe in the draft is
> stretching
> >>>>>>>> schema.org documentation style a little, by adding in an order
> >>>>>>>> of preference to the Expected Type.
> >>>>>>>>
> >>>>>>>> You are correct that a URI could be an identifier but the
> >>>>>>>> meaning I was hoping for in this case was that the identifier
> >>>>>>>> in question would be a [Standard] Identifier.  Hence the
> >>>>>>>> examples of ISSN, ISNI, etc.
> >>>>>>>>
> >>>>>>>> However I omitted the word 'standard' so as not to restrict
> the
> >>>>>>>> use to only identifiers produced by standards bodies.
> >>>>>>>>
> >>>>>>>> So, in my draft I was intending the URI would identify (link
> >>>>>>>> to) the description of a [standard] identifier.
> >>>>>>>>
> >>>>>>>> I perhaps need to work a little on my descriptive text - all
> >>>>>>>> suggestions welcome!
> >>>>>>>>
> >>>>>>>> ~Richard.
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> On 06/01/2013 18:01, "Karen Coyle" <kcoyle@kcoyle.net
> >>>>>>>> <mailto:kcoyle@kcoyle.net>> wrote:
> >>>>>>>>
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> On 1/6/13 8:27 AM, Gordon Dunsire wrote:
> >>>>>>>>>> Richard
> >>>>>>>>>>
> >>>>>>>>>> In the Identifier proposal, the value of the
> >>>>>>>>>> schema.identifier property includes URIs.
> >>>>>>>>>>
> >>>>>>>>>> 1. Assuming that schema.org Type is a synonym for RDF Class
> >>>>>>>>>> (or am I wrong?), what does this mean for the range of the
> >>>>>>>>>> identifier?
> >>>>>>>>>> Is
> >>>>>>>>>> it
> >>>>>>>> a
> >>>>>>>>>> literal, as in the "urn:nbn:ch:bel-9039" example, or a
> class,
> >>>>>>>>>> as in the <examplelib.org/identifier/12345> example, which
> is
> >>>>>>>>>> an
> >>>>>>>> individual
> >>>>>>>>>> member of the class?
> >>>>>>>>>
> >>>>>>>>> I believe that schema.org is purposely avoiding the strict
> >>>>>>>>> definition of domains and ranges. As it says in the
> documentation:
> >>>>>>>>>
> >>>>>>>>> "The decision to allow multiple domains and ranges was purely
> >>>>>>>> pragmatic.
> >>>>>>>>> While the computational properties of systems with a single
> >>>>>>>>> domain
> >>>>>>>> and
> >>>>>>>>> range are easier to understand, in practice, this forces the
> >>>>>>>>> creation of a lot of artifical types, which are there purely
> >>>>>>>>> to act as the domain/range of some properties. "
> >>>>>>>>>
> >>>>>>>>> http://schema.org/docs/datamodel.html

> >>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>> 2. Using the example, is it not true to say in ttl:
> >>>>>>>>>>
> >>>>>>>>>> <http://www.worldcat.org/oclc/38264520> schema.identifier
> >>>>>>>>>> <http://www.worldcat.org/oclc/38264520> .?
> >>>>>>>>>>
> >>>>>>>>>> If so, <http://www.worldcat.org/oclc/38264520> is both a
> >>>>>>>>>> schema.Book and a schema.Identifier Š
> >>>>>>>>>
> >>>>>>>>> If the two (schema.Book and schema.Identifier) are not
> >>>>>>>>> disjoint, does it matter?
> >>>>>>>>>
> >>>>>>>>> kc
> >>>>>>>>>>
> >>>>>>>>>> Cheers
> >>>>>>>>>>
> >>>>>>>>>> Gordon
> >>>>>>>>>>
> >>>>>>>>>> *From:*Richard Wallis[mailto:richard.wallis@oclc.org]
> >>>>>>>> <mailto:[mailto:richard.wallis@oclc.org]>
> >>>>>>>>>> *Sent:* 03 January 2013 12:36 *To:*public-schemabibex@w3.org
> >>>>>>>>>> <mailto:public-schemabibex@w3.org>
> >>>>>>>>>> *Cc:* Young,Jeff (OR); Vizine-Goetz,Diane
> >>>>>>>>>> *Subject:* Some Draft SchemaBibEx Proposals
> >>>>>>>>>>
> >>>>>>>>>> Hi all,
> >>>>>>>>>>
> >>>>>>>>>> I have pulled some of our thoughts and discussions together
> >>>>>>>>>> into a couple of draft vocabulary proposals.  They can be
> >>>>>>>>>> found on the Wiki
> >>>>>>>>>> here:
> >>>>>>>>
> >>>>>>>>
> <http://www.w3.org/community/schemabibex/wiki/Vocabulary_Proposals>.
> >>>>>>>>>>
> >>>>>>>>>> These are most definitely Œdraft¹ proposals and are there as
> >>>>>>>>>> a foundation for us to work on.
> >>>>>>>>>>
> >>>>>>>>>> I am not precious about any of the Type or Property names I
> >>>>>>>>>> have used, or any of the descriptive text either.  If you
> >>>>>>>>>> have better suggestions, dive in and share!
> >>>>>>>>>>
> >>>>>>>>>> I have included some example RDF ­ I will add some RDFa and
> >>>>>>>>>> possibly other format examples later.  I am holding off for
> a
> >>>>>>>>>> few days on this, as I am in discussion with the W3C hosting
> >>>>>>>>>> people about adding a syntax highlighting extension added to
> >>>>>>>>>> the Wiki which will make code examples far more readable.
> >>>>>>>>>>
> >>>>>>>>>> Regards,
> >>>>>>>>>>      Richard.
> >>>>>>
> >>>>>> --
> >>>>>> Karen Coyle
> >>>>>> kcoyle@kcoyle.net http://kcoyle.net

> >>>>>> ph: 1-510-540-7596
> >>>>>> m: 1-510-435-8234
> >>>>>> skype: kcoylenet
> >
> >
> >
>

Received on Wednesday, 16 January 2013 18:35:02 UTC